UNITED NATIONS CONFERENCE ON TRADE AND DEVELOPMENT
UNCTAD
ECONOMIC DEVELOPMENT IN AFRICA
RECLAIMING POLICY SPACE
Domestic Resource Mobilization
and Developmental States
UNITED NATIONS
ECONOMIC DEVELOPMENT IN AFRICA
UNCTAD/ALDC/AFRICA/2007
United Nations publication
Sales No. E.07.II.D.12
ISBN-13: 978-92-1-112723-2
ISSN 1990-5114
UNITED NATIONS
Printed at United Nations, Geneva
GE.07-51203–July 2007–6,730
Domestic Resource Mobilization and Developmental States
33
Chapter 2
THE CHaLLEnGE OF InCREaSInG SaVInGS
anD bOOSTInG PRODUCTIVE InVESTMEnT
A. Introduction
Achieving high and sustainable levels of investment needed for Africa’s
development will require a more balanced combination of foreign and domestic
resources than has been the case until now. It is expected that more reliance on
domestic resources will give countries more policy space to implement strategies
that reflect their development priorities, unlike past strategies, which were donor
driven. After all, African leaders are answerable to Africans, albeit imperfectly,
but donors are not. The resulting increase in ownership of the development
agenda should boost the efficiency of development strategies if they focus on
those sectors where investment is most productive.
Realistically, boosting domestic savings will require higher rates of economic
growth sustained over a long period. ODA could trigger this growth process
if it focused on financing pro-growth public investment such as economic
infrastructure. Such investment would in turn crowd in domestically financed
private investment.6 The higher growth rates resulting from these investments
would eventually generate more domestic resources and more investment,
sustaining high growth rates, and so on. Once set in motion, this virtuous circle
is possible even though Africans are thought to be too poor to save. After all,
poor countries in East Asia and other developing regions have been able to save,
invest and grow.
The relationship between savings and investment shows a mixed picture. In
the long term, the correlation between savings and investment rates is positive
and strong. From 1960 to 2005, the correlation coefficient between gross capital
formation and gross domestic savings was 0.74. In the short term, however, there
is almost no relationship between these two variables. In 2005, for example,
the cross-country correlation coefficient stood at only 0.05. Indeed, country
comparisons show that the countries with the highest investment rates are not
necessarily the ones with high savings rates. Sao Tome and Principe had the
34
Economic Development in Africa: Reclaiming Policy Space
highest rate of gross capital formation, representing 44 per cent of GDP, but
the second-lowest rate of gross domestic savings, representing –26.4 per cent of
GDP. Some of the highest investment rates in 2005 were recorded in countries
with the lowest savings rates. Zimbabwe, Lesotho, Ethiopia and Rwanda had
investment rates ranging from 38 per cent to 21 per cent of GDP against savings
rates of 3 per cent to –11 per cent of GDP (World Bank, 2007a).
One possible explanation of these contrasting results is that short-term
investments are strongly affected by external sources, particularly ODA, while
in the long term, investments are determined more by domestic savings.
Indeed, despite the positive role played by foreign aid in Africa’s development,
overdependence on ODA to fund domestic investments raises some issues.
Firstly, such investments are difficult to predict and sustain, given the high
volatility of ODA. Indeed, ODA to Africa has been found to be up to four times
more volatile than government revenue (UNCTAD, 2006a). Secondly, increases
in aid to specific countries have not usually been sustained for more than a
few years. Thirdly, the decision about which sectors to invest in may reflect
donor rather than national interests. The experience of aid conditionality in
the current poverty reduction strategy papers illustrates this point clearly (see
UNCTAD, 2002, 2006a).7 There is general consensus that Governments should
fully own their development programmes. Such “ownership” is difficult when
the implementation of the development programme is largely dependent on
external sources of finance, particularly when donors’ priorities do not coincide
with those of the recipient countries. Fourthly, aid cannot directly finance private
sector investment, which is the type of investment most needed to sustain rapid
growth in Africa. Furthermore, even when investment rates are high, efficiency
of such investments is sometimes low, reducing their potential positive effect on
economic growth (see chapter 3).
B. Increasing savings
In the light of the discussion in chapter 1 on the determinants of savings,
five propositions are explored in order to offer possible ways of increasing
savings in Africa: (a) controlling demographic factors; (b) financial sector reforms
and development of the informal financial sector; (c) increasing investment
opportunities; (d) reducing capital flight and inefficiencies in resource allocation;
and (e) reforming the tax sector.
Domestic Resource Mobilization and Developmental States
35
Controlling demographic factors
Africa is the only region of the world that has not achieved a demographic
transition (UNECA, 2005c). It has been argued that Africa’s high fertility rates8
have kept its dependency ratios very high.9 It is however, generally acknowledged
that high dependency ratios make savings difficult. As income is shared among
several people, very little is left for saving, particularly when household incomes
are relatively low. Therefore, for domestic savings to increase, the dependency
ratio should decline, per capita income should increase, or both. Decreasing the
dependency ratio is difficult in the short term. However, there are positive signs
that Africa may succeed in reducing its dependency ratio, as the continent’s
annual population growth rate is expected to decline from the current 2.5 per
cent to 2 per cent over the next 10 years (McKinley, 2005). Moreover, if the
relatively high economic growth rates achieved in Africa the last few years are
sustained, these rates, in combination with low population growth, could induce
higher savings rates.
These positive changes need to be balanced against the negative effect of
pandemics such as HIV/AIDS, malaria and tuberculosis on African countries’
capacity to save. These pandemics are destroying human capital at a rate never
experienced before. In some countries, particularly in Southern Africa, the rate of
attrition of professionals such as teachers, health workers, construction workers
and mineworkers is so fast that they cannot be replaced fast enough (UNECA,
2005c). In the absence of reliable social security systems, a growing number
of survivors are also using their resources, including their savings and time, to
provide care to the sick and orphans. As a result, these pandemics have dented
Africa’s meager financial and human resources, leading to even lower savings
rates.
Financial sector reforms and development
of the informal financial sector
Financial sector reforms carried out in Africa in the 1990s have had a
limited impact on domestic resource mobilization, as chapter 1 has shown.
The traditional concentration of the banking sector on import–export activities
to the detriment of productive investment in agriculture and industry has not
changed. Moreover, commercial banks are still concentrating their activities on
large clients, such as large firms in the private sector, as well as the public sector.
Furthermore, the fact that those banks have traditionally shied away from the
36
Economic Development in Africa: Reclaiming Policy Space
rural economy in Africa, where most productive activities are located, suggests
that important savings remain untapped. Indeed, the formal financial sector has
not adapted to the reality of the post-colonial State, where small and mediumsized enterprises dominate the economy. As Steel et al. (1997) put it, formal
banks have little interest and experience in serving small clients, particularly in
rural areas. Therefore, informal financial institutions should be empowered to
monetize and mobilize domestic savings, as they are more appropriately adapted
to developing countries’ financial needs.
There is some evidence that, although they had no important effect on formal
financial institutions, financial reforms remarkably increased the performance of
informal financial institutions (Steel et al., 1997). Survey evidence shows that
between 1990 and 1992 in the United Republic of Tanzania, the volume of
deposits in savings and credit cooperatives increased by 67 per cent, while
deposits in rotating savings and credit associations increased by 113 per cent,
thanks to higher numbers of clients and larger average savings.10
Similar trends were observed in Nigeria over the same period. Deposits in
savings and credit associations rose by 56 per cent and those of rotating savings
and credit associations by 77 per cent, while the capital base of savings and
loan companies grew by 148 per cent (Steel et al., 1997). In Burundi, although
microfinance is a relatively new phenomenon, its activities have increased
substantially over the last decade. Unlike the formal banking sector, microfinance
institutions are present in almost all of the country’s 116 administrative
communes. In 2005, the registered microfinance institutions had 300,000
members, or about 5 per cent of the population. They mobilized $14 million
and there is potential for further expansion (Nzobonimpa et al., 2006).
Despite its relevance and impressive growth, the microfinance sector by itself
cannot fill the current intermediation deficiencies. Microfinance institutions
have the advantage of covering a large number of small savers in areas where
formal banks are absent but the amounts involved in each operation are
relatively small, as the evidence from Burundi has shown. Moreover, as already
suggested, microfinance institutions are able to cover only a small fraction of
the total demand for financial services. The reasons for the limited expansion of
microfinance institutions and their inability to fill the intermediation gaps include:
(a) lack of stable resources, obliging the institutions to lend only in the short
term; (b) lack of human and technical capacity of the lenders and borrowers to
manage larger volumes of operations and design large investment projects; and
Domestic Resource Mobilization and Developmental States
37
(c) the high cost of credit due to high interest rates paid to commercial banks,
which provide the advances that microfinance institutions turn into loans to
members (see Nzobonimpa et al., 2006).
The message here is that even if microfinance institutions currently mobilize
relatively small amounts, they are able to provide financial services to segments
that the formal banking sector cannot reach. Moreover, the steady increase of
microfinance institutions over the years, particularly in comparison to the formal
banking sector, suggests that they have the potential to widen their services to
more clients traditionally outside the credit market. Hence, given their respective
roles, the formal banking sector and microfinance provide complementary credit
delivery mechanisms. Strengthening the synergies between these two types
of institutions would increase the quality and quantity of services provided to
the public. Chapter 4 proposes some specific strategies that could be used to
mobilize more stable domestic long‑term financial resources.
Increasing investment opportunities
Mainstream economists argue that the low rates of investment in developing
countries are primarily due to low savings. While this is true to some extent,
Keynesian economists point to the possibility of reverse causality, from low
investment to low savings due to the multiplier and accelerator effects.11 As
discussed earlier, the initial investment, funded by external resources, would
then create its own savings through the multiplier effect.
However, the mismatch in the term structure between savings (short-term)
and productive investments (long-term) implies that many countries have
problems transforming savings into productive investment, as illustrated by large
excess liquidity in the banking sector. A 2003 study of the banking sector in
the Central Africa region found that commercial banks maintained high liquidity
ratios. Liquidity ratios of banks in Cameroon, Chad and Equatorial Guinea were
higher than 200 per cent, while the ratios for Gabon and the Central African
Republic were 183 per cent and 101 per cent, respectively (Avom and Eyeffa,
2006). A similar situation was observed in Ghana, Malawi, Nigeria and the
United Republic of Tanzania in the 1990s (Steel et al., 1997) and Burundi in
2005 (Nzobonimpa et al., 2006).
As a result, only a small proportion of savings is allocated to long-term credit
that finances productive investment. In Burundi, for example, long-term credit
38
Economic Development in Africa: Reclaiming Policy Space
(more than five years of maturity) represented only 3 per cent of total credit
in 2004, compared with 17 per cent and 80 per cent for medium-term (one
to five years) and short-term (less than one year) credit, respectively. Although
this term structure is not particular to Africa, the absence of sophisticated bond
markets in the continent to provide reliable long-term savings and borrowing
instruments that limit the amount of long-term investible funds, create what
has been termed “original sin” facing most developing economies.12 Used in a
way that preserves a country’s domestic debt sustainability, bond markets not
only allow Governments to borrow in order to invest in long-term investment
projects, such as infrastructure, they also provide banks with an opportunity for
portfolio diversification.13 Therefore, the promotion of capital markets could
attenuate the problems due to the unavailability of long-term credit affecting
most countries in Africa.
Reducing capital flight and inefficiencies in resource allocation
The discussion in chapter 1 highlighted the extent to which capital flight
robs Africa of its much-needed financial resources. If these resources could
be efficiently invested within the continent, African countries could raise their
domestic savings remarkably. Capital flight is both a cause and a consequence
of a country’s poor investment performance. In the current context of high
capital mobility, capital flight may be a rational response to the lack of profitable
investment opportunities within Africa. Moreover, Boyce and Ndikumana (2001),
show that capital flight is intimately associated with a country’s external debt. In
this regard, debt relief could reduce capital flight and increase domestic savings,
particularly if complementary improvements in governance are made to ensure
that countries do not fall back into the debt trap.
Political and economic governance are also important determinants of capital
flight. Those involved in capital flight are usually Africa’s political and economic
elites, who are engaged in illegal practices to appropriate their countries’ wealth.
They use a variety of means, including trade misinvoicing; embezzlement of tax
revenue, exports and aid; and kickbacks on contracts. These practices result in a
skewed distribution of wealth within the society. On the one hand are members
of the elite who engage in capital flight and lavish spending rather than saving
and investing in their economies. On the other hand are poor people who have
to struggle to make ends meet. The result is a high level of income inequality,
illustrated by the fact that Africa has the world’s second-highest Gini coefficient
(42) after only Latin America (50).14 It is not by coincidence that countries such
Domestic Resource Mobilization and Developmental States
39
as Nigeria, South Africa and Zambia have a combination of some of the highest
levels of inequality and capital flight. These two issues are not just economic;
they have a governance dimension that needs to be addressed appropriately.
Inefficient spending is another channel through which African economies
are starved of the resources needed for productive investment. As it is difficult
to find a good measure of waste in the way resources are used, total factor
productivity growth could be used as a proxy measure of the efficiency with
which resources are used. Results based on macroeconomic accounting have
found that from 1960 to 2000, average total factor productivity growth in Africa
declined from a relatively high value of 2 per cent per annum in the 1960s (the
same as the world average) to negative values in the 1970–2000 period (Collins
and Bosworth, 2003). As a result, annual growth of output per worker in Africa
grew by only 0.6 per cent over the period, well below the world average of 2.3
per cent.
Although there are no statistics to support the argument that low total factor
productivity growth in Africa was due, at least partly, to allocative inefficiencies
or wasteful spending, political economy considerations could help explain why
resources are not optimally used. As noted in the case of capital flight discussed
above (see also the discussions in chapter 3), there have been instances where
state and private assets were appropriated by ruling elites who used them for
their own benefit. Collier and O’Connell (2007) use 26 detailed case studies
of growth in Africa over the period 1960–2000 to show, among other things,
how inefficient redistributions were associated with allocative inefficiencies. The
improvement in governance across the continent as illustrated by the adoption
of the African Peer Review Mechanism (see chapter 3 for details) has already
reduced the effect of many of these negative factors and improved the overall
image of the continent.
African leaders did not make these policy choices with the intention of
destroying their economies. In some instances, decisions that turned out
later to be inefficacious were made with the genuine belief that they would
lead countries to a path of growth and development (e.g. United Republic of
Tanzania President Julius Nyerere’s “African Socialism” or Ujamaa). One should
also add that these policies reflected the development paradigm at the time, and
were even supported by multilateral financial institutions (Mbabazi and Taylor,
2005: 7–8). Some other decisions, however, reflect institutional weaknesses that
failed to force some leaders to behave responsibly in the interest of the majority.
40
Economic Development in Africa: Reclaiming Policy Space
This was the case with some redistribution policies. These weaknesses require
modern and strong economic and political institutions that should outline the
rules defining the boundaries within which political leaders make their choices.
In other words, the State must be empowered to fully play its role in a context of
clearly defined obligations and prerogatives.
Reforming the tax sector
One area of economic governance that needs attention is the administration
of the tax system, considering that most Governments derive much of their
domestic revenue from taxation. If taxes were efficiently collected and
embezzlement tackled, tax revenue should increase substantially (Fjeldstad
and Semboja, 2001). Although the reason for uncollected taxes is partly due
to inefficiencies and limited administrative capacity, tax collectors often collude
with taxpayers in many countries to defraud the State of part of its revenue
(Fjeldstad, 2005).
A number of countries in Africa have changed their tax policies, sometimes
under pressure from their development partners. Initially, these reforms had
some success, as the Ugandan experience showed. The reforms helped create
tax systems that are more income elastic, more broadly based and less distorting
to economic activity. However, since it is much easier to change policies than to
change institutions, tax administration is still a major problem. Taxpayers do not
believe that their taxes are properly used to produce the expected basic services
the State should provide. As discussed in chapter 1, this reduces their incentive
to pay taxes, which could explain, at least partly, why tax policy reforms have so
far had limited effects (see Fjeldstad and Semboja, 2001).
Policies aimed at increasing government revenue through greater tax
compliance must balance three pillars. Firstly, the State must be legitimate
with a clear contract with taxpayers, where the obligations of the State towards
taxpayers are respected. For example, if the State invests the tax collected into
public and social services that benefit the community, taxpayers would be less
likely to engage in tax evasion as a form of protest against predatory authorities.
Secondly, tax collection entails some form of coercion. If it were left to each
individual to decide whether to pay taxes, opportunistic tax avoidance would
probably be rampant. Taxes are transfers from the fruit of individual or company
efforts to other entities, so it is the element of coercion that distinguishes them
from philanthropic donations. Thirdly, it is important to put in place a reliable
Domestic Resource Mobilization and Developmental States
41
mechanism of detecting and punishing tax evaders. Coercion would become
meaningless if tax evaders could not be detected and appropriately punished.
These three elements must be taken together. The relatively low tax compliance
in Africa seems to be due to the fact that countries have a tendency to put more
emphasis on the second pillar of the policy, without giving due regard to the first
and third pillars (Fjeldstad and Semboja, 2001).
C. Credit constraints
Credit is the main channel through which savings are transformed into
investments. Not all savings are used to finance investment, despite high
demand for credit, because the credit market in Africa is rationed. Indeed, the
lack of credit has been cited by firm managers in Africa as their most important
constraint (Bigsten and Soderbom, 2005). Access to credit is important because
it affects the level of investment made by a firm, which in turn is associated with
its growth and overall contribution to economic activity. Credit constraints also
affect the efficiency of investment (Bigsten et al., 1999). The lack of the required
amounts of credit can force a firm to postpone, scale down or even abandon
investment plans that are crucial for its economic viability, thus affecting the
firm’s profitability and growth. Moreover, financially constrained firms are less
able to adjust to short-term shocks to their cash flow, which could impair their
activities.
Recent studies have shown that the borrowing of firms from the banking
sector is very low in Africa. Based on a sample of manufacturing firms in Burundi,
Cameroon, Côte d’Ivoire, Ghana, Kenya and Zimbabwe, Bigsten et al. (2003)
found that 55 per cent of the firms did not apply for loans; 33 per cent were in
need of loans but their applications were rejected; and 12 per cent of the firms
received loans. Of course, the fact that a firm does not finance its activities with a
loan is not necessarily due to credit rationing, because some firms may not have
the need for credit. Credit rationed firms are those that apply for loans and have
their applications rejected, or those that do not apply because they believe their
applications would be rejected. As these figures show, a third of the firms in the
sample are credit rationed.
The same study found that the rate of credit rationing decreases as firm size
increases. For example, 64 per cent of firms with less than five workers and 42
per cent of firms with six to 25 workers are credit rationed. However, only 10
42
Economic Development in Africa: Reclaiming Policy Space
per cent of firms with more than 100 workers are credit rationed. About twothirds of the large firms do not participate in the credit market because they
use their retained profits. Considering that African economies are dominated by
small firms without adequate resources of their own, credit rationing is a critical
problem.
Why are so many firms and households unable to access credit even when
the banks have excess liquidity? Two major factors limit firms and households’
access to finance. Firstly, transaction costs associated with credit delivery to small
customers are high, particularly where large numbers of credit applicants live
in remote areas. The basic infrastructure essential for an efficient running of
banking operations is often lacking in rural areas, preventing formal banks from
penetrating this segment of the credit market. For example, it is inconceivable
for a bank to open branches in areas lacking key services such as access
roads, electricity or telephones,15 which are too costly to provide privately to
an atomized rural market. Secondly, information asymmetries between banks
and borrowers are high due to the absence of credit information systems. As a
result, banks are discouraged from extending their credit market because of the
difficulty of determining the true creditworthiness of borrowers and the high cost
of enforcing credit contracts once lending has taken place (Honohan and Beck,
2007).
High transaction costs
In most African countries, the oldest banks are financial institutions inherited
from the colonial period. These banks are often subsidiaries of metropolitan
parent banks, and their focus is on financing international trade and the services
sector, because these sectors provide quick high returns. Data collected in 2001
show that in Botswana, Guinea–Bissau and Lesotho, all banking assets were
owned by foreign banks. In addition, at least two thirds of banking assets were
owned by foreign banks in Benin, Côte d’Ivoire, Gambia, Guinea, Madagascar,
Mali, Namibia, Niger, Senegal and Swaziland (World Bank, 2007a).16 These banks
lend to larger borrowers such as the public sector, large enterprises and wealthy
households. They do not have mechanisms well suited to catering to the needs
of small, low-income and mostly agricultural and rural-based economic agents,
despite the fact that these agents constitute the backbone of African economies.
Instead, small borrowers have heavily relied on the informal financial sector.
Financial liberalization of the 1990s had little effect on this financial market
fragmentation.
Domestic Resource Mobilization and Developmental States
43
In contrast, microfinance institutions operate with minimum infrastructure,
making them more efficient in rural credit markets. Unfortunately, whereas
formal banks have excess liquidity, microcredit institutions have limited
resources, constraining their financial and geographic coverage. It is estimated
that microfinance institutions in Burundi were able to extend credit amounting
to about $16 million in 2005, satisfying only 30 per cent of the demand
(Nzobonimpa et al., 2006). Although it appears small, this coverage is much
higher than the rate of successful applications in the traditional banking sector
reported by Bigsten et al. (2003) for Burundi, Cameroon, Côte d’Ivoire, Ghana,
Kenya and Zimbabwe. Only 3 per cent of micro firms and 8 per cent of smallscale firms applied for and received credit. In addition, the estimated amount of
credit needed by those using microfinance institutions is 40 per cent higher than
total equity of all the country’s commercial and development banks. Hence,
those with relatively large investment projects cannot rely on microfinance to
meet their needs.
The second aspect of transaction costs incurred by formal banks relates to
their high administrative costs, which include the costs of screening, monitoring
and contract enforcement. In their study of the informal financial sector in African
countries in the early 1990s, Steel et al. (1997) find that traditional commercial
and development banks incur loan administrative costs that could be more than
10 times higher than microfinance institutions.17 Administrative costs of loans to
small-scale enterprises by commercial and development banks in Nigeria and
the United Republic of Tanzania represented 13 per cent and 12 per cent of
the loans, respectively. Comparatively, the costs were 1.9 per cent and 2.5 per
cent for credit unions in Nigeria and the United Republic of Tanzania, 1 per
cent and 0.1 per cent for savings and credit cooperatives, respectively. Urban
moneylenders’ administrative costs amounted to 3.2 per cent and 1.7 per cent
in Nigeria and the United Republic of Tanzania, respectively.
The foregoing shows that the financial market is segmented: the formal
banking sector focuses on large customers, whereas the informal sector
specializes in collecting savings from and lending to small customers. According
to Bigsten et al. (2003), a bank requires a profit to capital ratio of 200 per cent
to grant a loan to a firm with less than five permanent workers, all else being
equal, because it is too difficult and risky to deal with these small firms. The ratio
drops to 56 per cent if the firm has between 26 and 100 workers. Given the
dominance of small firms in African economies, appropriate measures should be
put in place to address their financial needs. For example, commercial banks’
44
Economic Development in Africa: Reclaiming Policy Space
excess liquidity could be efficiently used if there were proper channels through
which they could reach small and medium-sized enterprises that are in need.
Allowing communication between these two segments would be one way of
reducing the level of credit rationing affecting small firms.
Some countries have recently encouraged such communication through the
creation of semi-formal financial institutions or the establishment in commercial
banks of special windows dedicated to small borrowers. In South Africa, the
enactment of the Usury Act Exemption Notice in 1999 induced banks to create
their own micro-lending subsidiaries, which cater to the specific needs of small
borrowers previously kept outside the credit market.18 By 2002, these subsidiaries
accounted for almost half of the $2 billion owed to formal microcredit institutions
(Meagher and Wilkinson, 2002). In Central and West Africa, ECOBANK, a
regional commercial bank operating in 13 countries, has recently established
special windows dedicated to small and medium-sized enterprises.
Limited credit information
Credit trade in Africa is seriously handicapped by the lack of reasonable credit
information19 coverage, whether by private bureaus or public registries. The
only countries with good coverage of private bureaus are South Africa, where
the coverage rate in 2006 was 53 per cent of the adult population; Botswana
with 43 per cent; Swaziland with 39 per cent; and Namibia with 35 per cent.
Public registries have some importance in Cape Verde and Tunisia, each with 12
per cent of adults covered, followed by Benin and Mauritius, with 10 per cent
each. Sub-Saharan Africa has the second-lowest rate of private bureau coverage
in the world. On average, only 3.6 per cent of the adult population is covered, in
comparison to 60.8 per cent for OECD countries (World Bank, 2007c).
In the absence of credit information, banks tend to “personalize” credit,
concentrating their lending on a small group of clients who have developed
a reputation for creditworthiness through past interactions. Indeed, one of
the most interesting results emanating from the econometric literature on the
determinants of bank credit supply in Africa is that banks tend to lend to the
firms that are already in debt. The reason is that known creditworthy borrowers
are so few that banks compete to lend to them even before they finish paying
their previous debts. As a result, being in debt signals creditworthiness rather
Domestic Resource Mobilization and Developmental States
45
than a high risk of default. However, the relationship between debt and access
to credit is an inverted u-shape, implying a level of debt beyond which banks
are reluctant to provide more loans. Using econometric results in Bigsten et al.
(2003) where other determinants of credit are controlled for, this threshold can
be situated approximately at the level where debt represents 60 per cent of a
firm’s capital.
Ensuring creditworthiness ex ante is very important because collateral is not
a good protection against default. Because credit contracts are not properly
enforced, it is extremely difficult to seize and sell the collateral of a defaulting
borrower. A project with a high expected profitability rate signals the capacity but
not necessarily the willingness of the credit applicant to service his debt. In this
regard, credit contracts are implicitly based on the reputation of the borrower,
not on the value of collateral. The exceptionally high levels of collateral (up to
six times the amount of the loan) are used to indicate compliance with banking
regulations concerning portfolio risk profile (Nkurunziza, 2005b).
From the discussion above, two important policy conclusions emerge. Firstly,
financial policy reforms should acknowledge the existence and usefulness of a
segmented credit market and the important role played by microfinance. What
is thus needed is to put in place measures to create more synergies between
microfinance and the formal banking sector. Secondly, putting in place credit
information infrastructure and developing a legal system that enforces credit
contracts should be the two pillars of a new agenda for financial sector reforms.
An important caveat should be kept in mind in the discussion of credit
policies. Survey evidence shows that a large number of firms do not borrow,
either because they do not need credit or because they do not have the
capacity to repay. Extending credit across the board could harm such firms. For
example, an econometric study on the effect of credit on firm survival in Kenya’s
manufacturing sector in the 1990s, a period of economic turbulence, found
that firms using credit increased their odds of collapse by 92 per cent relative to
firms not using credit (Nkurunziza, 2005a). This was because these firms were
facing several other problems that eroded their capacities to use credit profitably.
Therefore, although the lack of access to credit is an important problem, it is just
one of many challenges facing firms in Africa.
Economic Development in Africa: Reclaiming Policy Space
46
D. Barriers to investment in Africa
High risk and a generally poor business environment are key determinants of
low investment rates in Africa. Due to credit rationing and high investment risk,
many economies in Africa are caught in a low-level “equilibrium of low demand
and low supply of credit” (Nzobonimpa et al., 2006: 17). As a result, firms’
decisions to invest are highly correlated with profitability. However, very high
profit rates are needed to convince an entrepreneur to make even some small
investments. For example, evidence based on survey data on manufacturing
firms in Cameroon, Ghana, Kenya and Zimbabwe shows that the firms that
invest are those with high profit rates, suggesting that firms rely on retaining
earnings to fund investment. However, due to the high risk prevailing in those
economies, only between 6 to 10 per cent of profits is invested (Bigsten et al.,
2003). The discussion in this section focuses on five main barriers to investment,
one structural and four institutional: (a) poor infrastructure, (b) high entry costs,
(c) labour market constraints, (d) low investor protection, and (e) high taxes
and a cumbersome tax system.20 Initial investment takes place when its riskadjusted expected benefits outweigh the costs. Table 2 gives a comparative view
of business costs in Africa relative to other regions.
Table 2
Regional comparative data on the cost of doing business
Region
Starting Licences Non-wage Credit Investor Total tax
Import
Export
Enforcing
business
cost
labour informa- protection
rate
cost
cost
contracts
(% GNI
(% per
cost
tion
index (% profit)
($ per
($ per
(% debt)
per cap.) cap. inc.) (% salary) index
container) container)
East Asia & Pacific
42.8
207.2
9.4
1.9
5.2
42.2
1,037.1
884.8
52.7
Europe & Central Asia
14.1
564.9
26.7
2.9
4.8
56.0
1,589.3 1,450.2
15.0
Latin America &
Caribbean
48.1
246.2
12.5
3.4
5.1
49.1
1,225.5 1,067.5
23.4
Middle East & North
Africa
74.5
499.9
15.6
2.4
4.6
40.8
1,182.8
923.9
17.7
882.6
OECD
South Asia
Sub-Saharan Africa
5.3
72.0
21.4
5.0
6.0
47.8
811.0
11.2
46.6
375.7
6.8
1.8
5.0
45.1
1,494.9 1,236.0
26.4
162.8
1,024.5
12.7
1.3
4.7
71.2
1,946.9 1,561.1
42.2
Source: World Bank (2007a).
Notes: Credit information index: ranges from zero to 6; high values mean more information availability.
Investor protection index: ranges from zero to 10; higher values mean better protection.
Measurement units of other variables are explained in their respective headings.
Domestic Resource Mobilization and Developmental States
47
Poor infrastructure
Poor infrastructure in many African countries discourages investment because
it increases production costs. This is the case with international transport costs,
defined here as the costs of moving goods between countries. As table 2 shows,
import and export costs are highest in Africa. On average, they represent 11 per
cent of the value of imports, compared with 7 per cent in Latin America and
the Caribbean, and 8 per cent in Asia. In landlocked African countries, transport
costs represent 19 per cent of import value (UNCTAD, 1999). These high costs
particularly inflate the cost of production of manufactured goods with a high
content of imported inputs. High international transport costs also increase
export costs, reducing exporters’ profit rates. As a result, high transport costs
have a significant effect on the location of foreign investors.
There is so far little microeconomic evidence on the impact of domestic
infrastructure on firm investment. However, opinion surveys of firm managers in
Africa report that poor infrastructure is the third-leading constraint to investment,
the other two being financing and corruption (Bigsten and Soderbom, 2005).
As explained earlier, the availability and reliability of basic infrastructure such
as access roads, electricity and telephones is essential to major investments,
so their absence or unreliability precludes such investments. In a number of
African countries, firms are obliged to invest in electricity generation by buying
their own generators, due to the unreliability of the national electricity grid. This
imposes a high cost and reduces the resources that could have been invested on
other productive activities. In many cases, potential investments are not made
due to the absence of basic infrastructure, which is complementary to other
investments.
High entry costs
Despite economic reforms undertaken since the 1980s, entry costs are still
relatively high in Africa.21 Survey data by the World Bank (2007a)22 show that in
2006, starting a business cost an average of 163 per cent of gross national income
(GNI) per capita in sub-Saharan Africa, the highest of all regions (see table 2).
Investing in Africa requires an average of 11 procedures, compared with only
six in OECD countries.23 In terms of the time needed to open a business, Latin
America and Caribbean countries have the longest waiting period (73 business
days), followed by sub-Saharan Africa with 62 days. The shortest waiting period is
in OECD countries, where on average only 17 days are required. These regional
48
Economic Development in Africa: Reclaiming Policy Space
averages hide important country variations. The cost of entry is lowest in South
Africa, where it represents 6 per cent of per capita GNI, followed by Mauritius
with 8 per cent and the Seychelles with 9 per cent.24 At the other extreme are
countries such as Sierra Leone, where the cost represents 1,194 per cent of GNI
per capita; the Democratic Republic of Congo, where it is 481 per cent; and
Niger, with a cost of 416 per cent.
High entry costs are deterrents to investment. In Africa, they are associated
with a small formal private sector and a large public sector, measured in terms of
formal employment. Before the introduction of SAPs in the 1980s, modern sector
employment was concentrated in the public sector in many African countries. In
Guinea, for example, state-owned enterprises alone accounted for 76 per cent
of formal employment in 1984. The corresponding proportion was 44 per cent
in Ghana in 1991, 43 per cent in Zambia in 1980, and 36 per cent in Burundi
in 1988 (World Bank, 2005a). However, as SAPs forced African countries to
trim their bureaucracies, the resulting “redundant labour” was unable to enter
into the formal private sector, partly due to the difficulties associated with entry
discussed above. Instead, many of those affected were forced to enter into the
informal economy to earn a living. Hence, retrenchment in the public sector had
the effect of increasing the “informalization” of economic activities.
Kenya offers a good illustration of how this unfolded in the 1980s and 1990s.
The need to reduce budget deficits led to a decline in public sector employment
from 30 per cent of total employment in 1990 to 11 per cent in 2000. This
“redundant labour” did not migrate to the modern private sector, since the latter
also shrank from 36.5 per cent of total employment to 17 per cent over the
same period. However, as the formal economy was losing ground, the informal
sector thrived. Employment in the non-agricultural informal sector tripled, from
about 23 per cent of total employment in 1990 to 70 per cent in 2000 (UNECA,
2005c).
While there are legitimate reasons for controlling entry, for example, in
situations where uncontrolled entry could harm consumers’ welfare, it is
important to recognize that unduly high entry costs as experienced in Africa limit
investments and hence their expected welfare effects. Therefore, controls need
to be balanced with the need to increase productive investment. Although entry
costs are currently high, they can be reduced substantially in a short period. For
example, Equatorial Guinea slashed its cost of entry from 2,051 per cent of GNI
per capita to 101 per cent in one year between 2005 and 2006. In Ethiopia, the
Domestic Resource Mobilization and Developmental States
49
waiting period to start a business was reduced by almost two thirds, from 44 to
16 days between 2003 and 2006 (World Bank, 2007c). This remarkable progress
was possible because the respective Governments were willing to reduce these
entry costs through a simplification of administrative procedures.
Labour market constraints
There is no consensus among economists on the net benefits of labour
regulations. Proponents are of the view that these are essential to ensuring
the safety, basic rights and fair compensation of workers. On the other hand,
advocates of a flexible labour market argue that these regulations increase
the cost of labour, which discourages job creation. Hence, they contend,
employment policies must be flexible enough to allow firms to adjust their
activities to economic shocks (UNECA, 2005c). Both views are founded. The
right balance between protection and flexibility delivers maximum benefits to an
economy. A “rigidity of employment index (REI)” is used to empirically assess the
situation in Africa (World Bank, 2007c).25
Sub-Saharan Africa has an REI of 47 out of 100. In comparison, the REI for
OECD countries is 33 and that of the East Asia and Pacific region is 23. There are
wide variations across countries. The REI is as low as 7 in Uganda and as high
as 78 in the Democratic Republic of the Congo. However, the REI gives only a
partial picture of the cost of employment regulations. Non-wage labour costs,
which include social security payments and firing costs, are other important
components of labour regulations. Sub-Saharan Africa’s non-wage labour costs
represent about 13 per cent of salary, the third lowest of all regions (see table
2).
Southern African countries, with the exception of Zambia, have very low costs,
representing less than 5 per cent of salary. On the other hand, francophone West
African countries have among the highest non-wage labour costs, representing
29 per cent of salary in Benin and Congo.26
Firing costs, expressed as the number of weeks of wages, represent the cost
of advance notice requirements, severance payments and penalties due when
terminating a worker. They are highest in sub-Saharan Africa and South Asia,
where they represent 71 weeks of wages, compared with 26 weeks in Europe
and Central Asia, where they are lowest. Firing costs range from nine to 13
weeks of wages in Gambia and Uganda, respectively, to a staggering 446 and
50
Economic Development in Africa: Reclaiming Policy Space
329 weeks of wages in Zimbabwe and Sierra Leone, respectively. While it is
reasonable that workers should be protected against abusive dismissal, requiring
a firm to pay almost nine years of salary in case it dismisses a worker, as is the
case in Zimbabwe, deters potential employers from hiring workers in the first
place.
What lessons may be drawn from the statistics above? Although Africa appears
to have put in place costly labour regulations, they have a small average effect on
employment and workers’ protection, but a large marginal effect on job creation
(UNECA, 2005c). The average effect is small because these regulations are only
limited to the formal sector, which accounts for a tiny share of the labour force. For
example, in the countries forming the Communauté Financière d’Afrique, only
about 5 per cent of the labour force is effectively covered by labour regulations,
which seem inflexible. However, even when the regulations are in place, they
often are not enforced, but this may not be known before entry by potential
investors. As a result, few workers are effectively protected by labour regulations.
Furthermore, some of these regulations are not binding. For example, minimum
wages are set at levels below actual starting wages in many countries, such as
in Ghana’s and Kenya’s manufacturing sector. The disconnect between written
regulations and their implementation may explain why empirical studies do not
find a statistically significant relationship between hiring and firing rules and the
level of unemployment in developing countries.
The fact that the average effect of these regulations on labour markets is small
does not mean that they are not important. Whether or not they are enforced,
regulations have a strong signalling effect. A potential entrepreneur might decide
not to invest if he perceives these regulations as signalling an unfavourable
investment environment. Foreign investors, particularly, are most sensitive to the
signalling effect of labour regulations for two main reasons. Firstly, they are less
able to know that the regulations are not enforced in the domestic economies.
Secondly, foreign capital is more mobile, implying that a small difference in
labour costs may have a large impact on the location of foreign investment. In
this light, the perception of rigidity of African labour markets may help explain
why the low level of foreign investment flowing to Africa targets mostly strategic
sectors such as extractive industries, with little impact on employment creation
(UNECA, 2005c).
The decision facing domestic investors is whether to invest in the formal or
informal sector. If they perceive labour regulations in the formal sector to be
Domestic Resource Mobilization and Developmental States
51
excessively cumbersome and costly, they might decide to invest in the informal
sector. For example, a recent study on India found that States that amended
labour regulations in favour of workers experienced a decline in output,
employment, investment and productivity in formal manufacturing firms, while
output increased in informal manufacturing firms (Besley and Burgess, 2004).
Therefore, as for entry costs, labour market regulations should balance the need
for workers’ protection against the need for investment. After all, it is only by
allowing investment to take place that jobs can be created and workers can
enjoy their work-related benefits.
Low labour productivity is another important deterrent of investment in
Africa, particularly FDI. It has been established that African firms are generally
less productive than their counterparts in other developing and developed
economies (see Bigsten and Soderbom, 2005). The consequence is that firms try
to minimize the amount of labour they hire, even if there is high unemployment
and wage rates are low. This prevents the growth of a dynamic labour-intensive
manufacturing sector, as has occurred in East Asia. Even countries that have
offered foreign investors generous incentives, such as access to export promotion
zones, have failed to attract much labour-intensive investment.
Low investor protection
Investor protection can be understood in two different but complementary
ways. The first refers to: (a) the extent of disclosure to ensure transparency of
transactions; (b) the extent of liability of a firm manager for self-dealing; and
(c) the ability of shareholders to sue a company’s officers and managers for
misconduct (see World Bank, 2007c).27 Based on these elements, there are
no important differences in investor protection between SSA and other regions
(see table 2). On a scale of zero to 10, where higher values represent higher
protection, the index of investment protection in South Africa and Mauritius is
8 each, the highest in Africa. At the other extreme are Djibouti and Swaziland,
each with a score of 2. Botswana, a country hailed for its economic success, has
a score of 4, which is below the African average of 4.7 per cent.
One of the most relevant aspects of investor protection in Africa is the
vulnerability of firms to arbitrary decisions by public officials, irrespective of the
prevailing formal laws and regulations, which can be very costly. For example,
Gauthier and Gersovitz (1997) found that medium-sized firms in Cameroon
are penalized, as they appear to be the only ones paying taxes. The reason is
52
Economic Development in Africa: Reclaiming Policy Space
that small firms in the informal sector could easily evade taxation, if necessary,
by shutting down and opening up elsewhere. On the other hand, large firms
have enough resources to lobby government officials for tax exemptions. The
consequence is that formal investment tends to be confined to large firms, while
small firms operate in the informal sector. This could help explain the so-called
“missing middle” phenomenon characterizing many developing economies’
industrial organization.
Investor protection may also be understood in terms of contract enforcement,
the most relevant form of protection in Africa. In the case of a payment dispute
over a claim amounting to 200 per cent of GNI per capita in sub-Saharan Africa,
it takes on average 581 business days and 38 procedures to settle the case, at a
cost representing 42 per cent of the debt. Comparing figures in the rest of the
world, South Asia has the longest waiting period, 969 business days, but the
number of procedures is similar to that in Africa, and the cost is lower, at 37
per cent of the debt. In Africa, Uganda has the most simplified system, with 19
procedures. The shortest waiting time in Africa is in Gambia, with 247 business
days. The cost of settling a dispute is lowest in Algeria and Gabon, where it
amounts to 10 per cent of the debt.
The main message is that Africa has among the lowest levels of investment
protection. In countries such as the Democratic Republic of the Congo, Malawi,
Mozambique and Sierra Leone, settling a business dispute could cost up to twice
the amount of the debt. As a result, businesses do not rely on the formal legal
system to enforce their commercial contracts. Most prefer to settle their disputes
amicably (Fafchamps, 1996).28 One problem with informal contract enforcement
mechanisms is they are unpredictable.
High taxes and a cumbersome tax system
High taxes are often blamed for the low level of investment in African
economies. A typical firm in sub-Saharan Africa pays the equivalent of 71 per
cent of its profits in taxes, which is 15 percent higher than the second-highest
rate, paid in Europe and Central Asia.29 In countries such as Burundi, the Central
African Republic, the Democratic Republic of the Congo, Gambia, Mauritania
and Sierra Leone, the amounts of taxes paid by firms are much higher than
their net profits. The justification that such poor countries must collect maximum
revenue with high tax rates to fund public services and reduce fiscal deficits is
challenged by a recent study by McLiesh and Ramalho (2006), which shows that
Domestic Resource Mobilization and Developmental States
53
high tax rates lead to low revenue, as they drive firms into the informal sector,
reducing the tax base. The study also shows that in poor countries, if the average
tax rate paid by businesses is 10 per cent, it would yield revenue equivalent to
16 per cent of GDP; if the tax is increased to 90 per cent, it would yield only
about 12 per cent of GDP in tax revenue (McLiesh and Ramalho, 2006: figure
2.1).
Tax payment also has a high bureaucratic cost, understood here as the
number of payments required each year. Fifty payments are required in Europe
and Central Asia, the highest of all regions, in comparison to 41 payments in subSaharan Africa, where making these payments takes 336 hours. These costs are
high relative to OECD countries, where 15 payments take 203 working hours.
Seychelles has the best record, with firms there making 15 payments in only 76
hours. In some countries, firms are required to spend so much time dealing with
tax payments that it becomes preferable to evade taxation.
There are two main messages arising from this information. Firstly, high tax
rates and costly tax procedures may discourage investors from investing in Africa’s
formal sector. Secondly, this tax system encourages tax evasion and capital flight,
reducing even further the financial resources available for development. This
does not imply, however, that African countries should engage in a “race to the
bottom” by competing to attract foreign investors through unreasonably high
tax cuts and tax breaks, as observed in many cases of FDI in natural resource
projects.30 In fact, given the importance of tax revenue for public investment,
the amount of expected tax from a foreign investment should be one of the
key determinants of its economic benefit to the country. Hence, as in the case
for entry costs and labour regulations, what countries need to put in place is a
simplified and predictable tax system that balances the profit-making interests of
investors and revenue generation for the host country.
E. Effect of business environment on
gross domestic capital formation
As expected, the difficult business environment in Africa appears to have
negatively affected investment. The correlation between a country’s rank in
terms of its ease of doing business and the rate of fixed capital formation in 2005
was –0.33, which is statistically significant. This negative relationship is illustrated
in figure 5.31
Economic Development in Africa: Reclaiming Policy Space
54
The focus on domestic economic factors in analyzing this negative relationship
stems from the recognition that they are important, but this in no way suggests
that external factors are less important. In fact, as chapter 3 argues, Africa’s low
savings and investment rates are due to a number of economic and non-economic
factors, both internal and external. Nevertheless, the focus on domestic factors
is due to the fact that, while it is difficult to alter external factors in favour of
Africa, it is relatively less costly to improve the domestic environment. There are
several administrative measures, some of them discussed in chapter 3, which
could substantially improve domestic savings and their efficient investment. For
example, creating an environment where formal and microfinance institutions
work in symbiosis has the potential to increase savings and improve credit
allocation. Moreover, increasing the efficiency of tax collection trough some
basic training and better monitoring of tax collectors could double the tax
revenue collected by the Government. Reducing tax exemptions could have an
additional positive effect on tax revenue.
Figure 5
Capital formation vs. business environment
32
Capital formation as % of GDP
28
24
20
16
12
8
4
0
40
80
120
Ease of doing business
160
200
Source: Based on data from World Bank, 2007a and World Bank, 2007c.
Note: Higher ranks on “Ease of doing business” represent poorer business environment.
Domestic Resource Mobilization and Developmental States
55
With respect to the investment climate, it is encouraging that most of the
barriers discussed in the chapter can be reduced relatively easily. Indeed, the
examples of Equatorial Guinea and Ethiopia have shown how the cost of entry
was reduced drastically in a short period. Sudan also reduced its total tax rate
from 54 per cent to 37 per cent of profit between 2005 and 2006. Many easy
steps can be taken to improve Africa’s business environment. For example,
unnecessary regulations that increase business costs or send negative signals to
investors without serving any purpose should be abandoned. This is the case with
some entry requirements and some labour regulations. Indeed, there is ample
room for simplifying business regulations simply through the promulgation of
appropriate administrative measures.
However, despite the negative correlation between the business environment
and capital formation, it would be naïve to expect that reducing the barriers
to investment discussed above would lead by itself to an investment boom in
Africa. A favourable business environment is not sufficient for achieving increased
investment rates. This suggests that, no matter how important these barriers to
investment discussed above are, there are other factors, particularly external and
political factors, which must be addressed in order to find long-lasting solutions
to Africa’s low level of savings and investment. Also important to Africa’s overall
trade and development performance are the global economic environment and
multilateral trade rules.
What emerges from the analysis is that action to change the situation in Africa
has to come from national Governments, although Africa’s development partners
also have a role to play to ensure that the global trading environment is not too
inimical to Africa’s development interests. This notwithstanding, most of the issues
discussed above require the State to take the initiative to introduce the measures
needed for mobilizing more resources and investing them more efficiently.
For example, only the Government can make decisions on the simplification
of business procedures. Moreover, some of the problems identified are due to
market failures that can only be addressed by government intervention. This is
the case, for example, with the provision of law and order, efficient mechanisms
for contract enforcement, and basic infrastructure such as roads and electricity,
etc., which constitute an essential basis for private investment. In other words,
the Government must assume the responsibilities of a developmental State to
help Africa emerge from its economic stagnation. Chapter 3 discusses in more
detail the argument in favour of developmental States in Africa.
56
Economic Development in Africa: Reclaiming Policy Space
Institutional reform will necessarily form part of any successful development
strategy in African countries. It is important, however, to emphasize that there is
no set blueprint for an institutional design that promotes development. Indeed,
both high‑income countries and fast‑developing countries have achieved their
highest growth rates within an institutional context that was markedly different
from that currently advocated by Bretton Woods Institutions and other donors
(UNCTAD, 2002). In fact, it has even been suggested that the imposition of
institutions that reflect current “best practice” in developed countries, rather
than those that have promoted their growth, represents a case of rich countries
“kicking away the ladder” with which they themselves climbed to success
(Chang, 2002).